This project involves the prediction of the three-dimensional structure of an octapeptide implicated in breast cancer, in that it elicits the same T-cell response as a surface glycoprotein found on cancerous cells. The sequence of the octapeptide is the same as the N-terminal region of the surface glycoprotein. The goal of this work is to find an analog which elicits a strong T-cell response to the glycoprotein on cancerous cells. The difficulty with the octapeptide is that it is quite hydrophobic (insoluble) and is easily metabolized, thus necessitating the use of an analog. A novel low energy structure was found for the octapeptide after a week of simulation on our Star Technologies ST-100 array processor (the equivalent of 3 years of computing on a VAX 11/780). The structure is supported by the experimental data available to date. From this work, several follow-up experiments have been suggested to test the structure. We are starting new simulations on related peptides that are about 10 residues in length.